Abstract: An apparatus includes: one or more processors configured to: randomly split a training data set into a first training data set comprising a first label assigned to first data and a second training data set comprising a second label assigned to second data; train a first neural network using a semi-supervised learning scheme based on the first training data set comprising the first label, and an unlabeled second training data set; and train a second neural network using the semi-supervised learning scheme based on the second training data set comprising the second label, and an unlabeled first training data set.

Abstract: A method includes receiving first data elements of a first data type and second data elements of a second data type. The first data type is text data and the second data type is at least one of image data or video data. The method also includes identifying first features of each of the first data elements, identifying second features of each of the second data elements, and generating merged features by combining a first feature of the first features of each of the first data elements with a second feature of the second features of one of the second data elements. The first feature and the second feature each represent a common feature. The method also includes generating a model based on the common features and at least a portion of the first features and the second features.

Abstract: Methods and systems for smoothening the transition of reward systems or datasets for actor-critic reinforcement learning models. A reinforcement model such as an actor-critic model is trained on a first dataset and a first reward system. The weights of the actor model and the critic model are frozen. While these weights are frozen, an affine transformation layer is attached to a final layer of the critic model, and the affine transformation layer is trained with a second dataset and a second reward system in order to adjust a weight of the final layer of the critic model. Then, the weights of the critic model are unfrozen which allows the adjusted weight of the final layer of the critic model to be implemented. The reinforcement learning model is retrained on the second dataset and second reward system, first with just the critic weights unfrozen, and then with both actor and critic weights unfrozen.

Abstract: An information processing device includes a feature extraction means for extracting, from input data that is motion data representing a motion of a person, basic feature data representing a feature of the motion data corresponding to a basic motion set with respect to the motion, motion feature data representing a feature of the motion data corresponding to a motion style set with respect to the motion, and person feature data representing a feature of the motion data corresponding to the person; a motion data generation means for generating first motion data based on the basic feature data and the motion feature data, and generating second motion data based on the basic feature data and the person feature data; and a learning means for learning the feature extraction means and the motion data generation means based on the first motion data and the second motion data.

Abstract: A computer-implemented method, according to one approach, includes issuing a hyperparameter optimization (HPO) query to a plurality of computing devices. HPO results are received from the plurality of computing devices, and the HPO results include a set of hyperparameter (HP)/rank value pairs. The method further includes computing, based on the set of HP/rank value pairs, a global set of HPs from the HPO results for federated learning (FL) training. An indication of the global set of HPs is output to the plurality of computing devices. A computer program product, according to another approach, includes a computer readable storage medium having program instructions embodied therewith. The program instructions are readable and/or executable by a computer to cause the computer to perform the foregoing method.

Abstract: A method, a computer program product, and a system of personalized training a machine learning model using federated learning with gradient boosted trees. The method includes training a global machine learning model using federated learning between a plurality of parties. The method also includes distributing the global machine learning model to each of the parties and receiving personalized model updates from each of the parties. The personalized model updates are generated from updated models boosted locally and produced by each of the parties using their respective local data. The method further includes fusing the personalized model updates to produce a boosted decision tree to update the global machine learning model. The method also includes training global machine learning model, iteratively, in this manner until a stopping criterion is achieved.

Abstract: A method for training a machine learning model is described, comprising receiving, for each perturbation of a plurality of perturbations of model parameters of a starting version of the machine learning model, a change of loss of the machine learning model caused by the perturbation for a set of training data determined by feeding the set of training data to one or more perturbed versions of the machine learning model, estimating a gradient of the loss of the machine learning model with respect to the model parameters from the determined changes of loss and updating the starting version of the machine learning model to an updated version of the machine learning model by changing the model parameters in a direction for which the estimated gradient indicates a reduction of loss.

Abstract: Methods, apparatus, systems, and articles of manufacture to modify pre-trained models to apply neural architecture search are disclosed. Example instructions, when executed, cause processor circuitry to at least access a pre-trained machine learning model, create a super-network based on the pre-trained machine learning model, create a plurality of subnetworks based on the super-network, and search the plurality of subnetworks to select a subnetwork.

Abstract: Implementations of the specification provide a knowledge graph data fusion method and system, and the method includes: obtaining a target entity field and a target relationship description, the target entity field and the target relationship description being selected from ontology definition data of two or more knowledge graphs; and then, obtaining data instances of related platforms or technology fields, and processing the obtained data instances based on a graph operator that is in ontology definition data of a fused knowledge graph and that is used to perform fusion processing on entity fields and relationship descriptions of different platforms or technology fields, to generate the fused knowledge graph.

Abstract: This specification relates to the field of knowledge graphs, and in particular, to knowledge reuse methods, apparatuses, computer-readable media, and systems. In an example computer-implemented method, a child entity is defined based on a parent entity, where the parent entity is selected from entities in a knowledge graph. A knowledge inheritance method is performed to inherit a portion of instance data of the parent entity. Instance data of the child entity are determined. First graph increment information corresponding to the instance data of the child entity is stored.

Abstract: A system coordinates services between users and providers. The system trains a computer model to predict a user state of a user using data about past services. The prediction is based on data associated with a request submitted by a user. Request data can include current data about the user's behavior and information about the service that is independent of the particular user behavior or characteristics. The user behavior may be compared against the user's prior behavior to determine differences in the user behavior for this request and normal behavior of prior requests. The system can alter the parameters of a service based on the prediction about the state of the user requesting the service.

Abstract: In an aspect an apparatus for generating emissions predictions is presented. An apparatus includes at least a processor and a memory communicatively connected to the at least a processor. A memory contains instructions configuring at least a processor to receive transport data of a transport from at least a transport entity. At least a processor is configured to extract emission data from transport data. At least a processor is configured to classify emission data to a transparency level. At least a processor is configured to generate an emissions prediction as a function of emission data and a transparency level. At least a processor is configured to display, through a graphical user interface, an emissions prediction and a transparency level to a user.

Abstract: A network-based service is provided that utilizes a machine learning model (MLM) trained on a variety of data from disparate systems of a retailer to generate predictive guidance/parameters for a price markdown process. The predictive guidance includes a markdown prediction that indicates whether an item should or should not be marked down. For an item designated for markdown, the MLM also generates markdown parameters including a markdown level for the item and a quantity of the item to be marked down. The predictions of the MLM are optimized to reduce item shrink, reduce item spoilage, increase item sales, and increase item margins. The service can be integrated into existing retailer systems and services to provide optimal markdown instructions for perishable items that are data-driven and objective.

Abstract: A method finds short counterfactuals. The method includes receiving an input vector with a plurality of input features. The method further includes processing, with a model, the input vector to generate a score. The score of the input vector is not to a selected class. The method further includes searching for a counterfactual vector using a cost value and a heuristic value. The searching includes replacing one or more input features of the input vector with one or more counterfactual features to generate the counterfactual vector. The counterfactual vector corresponds to a counterfactual score to the selected class. The method further includes presenting one or more recommendations using the counterfactual vector.

Abstract: Various embodiments for a continuous learning system are described herein. An embodiment operates by receiving a query from a user and identifying an unknown phrase in the query. User feedback regarding the unknown phrase is requested and received. A first pre-existing entity of a plurality of pre-existing entities that corresponds to the received user feedback is identified. A relationship between the first pre-existing entity and the unknown is added to the knowledgebase. The query is executed against the knowledgebase using the first pre-existing entity. A response, to the executed query, is provided to the user.

Abstract: The present invention relates to a method, a service server, and a computer-readable medium for selectively extracting data for labeling, and more particularly, to a method, a service server, and a computer-readable medium for selectively extracting data for labeling, capable of selectively providing data for effectively training a machine learning model that is desired to be achieved by a user by deriving a plurality of feature vectors for a plurality of data included in an original dataset, applying a preset rule to the feature vectors to select a preset number of feature vectors among the feature vectors, and plotting the selected feature vectors and unselected feature vectors on a plane of three or less dimensions to provide a plotting result to the user.

Abstract: One or more systems, devices, computer program products and/or computer-implemented methods of use provided herein relate to a process to facilitate abnormal document self-discovery. A system can comprise a memory that stores computer executable components, and a processor that executes the computer executable components stored in the memory, wherein the computer executable components can comprise an object detection component that can generate a knowledge graph with vectors corresponding with nodes representative of a page layout of a document. Additionally, the computer executable components can comprise an evaluation component that can compare the vectors to identify whether an edge is present between corresponding vectors of the knowledge graph; an encoder component that can re-code the knowledge graph; and a comparison component that can compare a structure of the knowledge graph with one or more other knowledge graphs corresponding to one or more other documents to determine if the document is abnormal.

Abstract: A candidate virtual team is generated using a team prediction model. The team prediction model is trained based on user feedback. Activity data of a user is collected, the activity data is associated with activities in which the user and a set of other users have engaged. A candidate virtual team associated with the collected activity data is generated using a team prediction model and the candidate virtual team is presented to the user using an interface. User feedback data associated with the generated candidate virtual team is received from the user. The team prediction model is trained using the received user feedback data, whereby accuracy of future generated candidate virtual teams generated by the team prediction model is improved during the training. The training of the team prediction model based on user feedback improves the generation of virtual teams over time.

Abstract: A method for training a machine-learning module of a computer-implemented prediction model for predicting product quality parameter values for one or more quality parameters of a chemical product produced by a chemical production plant. The production plant includes a plurality of sensors, each of which is configured to acquire process parameter values for one or more process parameters of a chemical process carried out by the production plant for producing the chemical product during operation of the production plant. A priori information about the production plant and the process carried out by the production plant is used, including chronological sequence information about a chronological sequence of the process carried out within the production plant, for which sensors sensor-specific time shifts between an acquisition time of training process parameter values and a production time of a product unit, during the production of which the corresponding training process parameter value was acquired.

Abstract: A system receives a predictive model and receives one or more runtime constraints. The system generates a directed acyclic graph (DAG) of the predictive model indicating dependencies. The system compiles the predictive model into first instructions for a first processor based on the one or more runtime constraints and the DAG. The system packages first instructions, the one or more runtime constraints, and the DAG of the predictive model in a first binary. The system recompiles the predictive model into second instructions for a second processor based on the runtime constraints and the DAG stored in the first processor. The system packages the second instructions, the DAG, and the runtime constraints in a second binary.

Abstract: A computer performs a first step of acquiring initial structures that are atomic arrangement structures in a three-dimensional space which a composition of a material can take, a second step of calculating first energy corresponding to structurally optimized atomic arrangement structures by performing structure optimization on one or some initial structures, a third step of predicting second energy corresponding to atomic arrangement structures obtained in a case where structure optimization is performed on other initial structure(s) by using a prediction model for the other initial structure(s), a fourth step of extracting third energy indicative of a minimum value on the basis of the first energy and the second energy, and a fifth step of outputting the third energy, a first structure, which is an atomic arrangement structure corresponding to the third energy, or the third energy and the first structure.

Abstract: A computer performs a first step of acquiring initial structures that are atomic arrangement structures in a three-dimensional space which a composition of a material after desorption of an atom contained in the material can take, a second step of calculating first energy corresponding to structurally optimized atomic arrangement structures by performing structure optimization on one or some initial structures, a third step of predicting second energy corresponding to atomic arrangement structures obtained in a case where structure optimization is performed on other initial structure(s) by using a prediction model for the other initial structure(s), a fourth step of extracting third energy indicative of a minimum value on the basis of the first energy and the second energy, and a fifth step of outputting the third energy, a first structure, which is an atomic arrangement structure corresponding to the third energy, or the third energy and the first structure.

Abstract: A method includes determining, by a processing device, a criterion associated with a configuration of a containerized computing cluster, wherein the containerized computing cluster comprises a plurality of virtualized computing environments running on one or more host computer systems; evaluating a rule against a fact, wherein the rule specifies a condition including the criterion and an action to perform if the condition of the rule is satisfied; and responsive to determining that the condition specified by the rule matches the asserted fact, performing the action specified by the rule, wherein the action comprises a notification regarding the configuration of the containerized computing cluster.

Abstract: A method for computer question answering includes, at a retriever subsystem of a question answering computer system, identifying a plurality of relevant text evidence strings for an input text question. At a linker subsystem of the question answering computer system, one or more of the plurality of relevant text evidence strings are associated with a respective secondary text evidence string to form a plurality of evidence chains via a previously-trained entity-linking machine-learning model. At a chainer subsystem of the question answering computer system, a ranked set of the evidence chains is identified based at least in part on an output of a generative machine-learning model applied to each of the plurality of evidence chains. At a reader subsystem of the question answering computer system, an answer to the input text question is output based at least in part on the ranked set of evidence chains.

Abstract: A two-stage machine learning model is used to for categorization of a dataset, such as transactions. A plurality of complementary base machine learning models are used to generate initial inference results and associated measures of inference confidence from the dataset, which are collected as a meta dataset. Each of the complementary models is associated with a different part of the dataset in which it has a higher accuracy in that part than the other models. The meta dataset is provided as input to a meta machine learning model, which is trained to produce a final inference result, and a confidence score model, which is trained to produce a confidence score associated with the final inference result.

Abstract: This document relates to automated generation of machine learning models, such as neural networks. One example method involves obtaining a first machine learning model having one or more first inference operations. The example method also involves identifying a plurality of second inference operations that are supported by an inference hardware architecture. The example method also involves generating second machine learning models by modifying the first machine learning model to include individual second inference operations that are supported by the inference hardware architecture. The example method also involves selecting a final machine learning model from the second machine learning models based on one or more metrics.

Abstract: A maintenance solution pipeline is automatically selected from a plurality of maintenance solution pipelines, based on obtained information. The maintenance solution pipeline is to be used in providing a physical asset maintenance solution for a plurality of physical assets. Code and model rendering for the maintenance solution pipeline automatically selected is initiated. Output from an artificial intelligence process is obtained. The output includes an automatically generated risk estimation relating to one or more conditions of at least one physical asset of the plurality of physical assets. Code and model rendering for the maintenance solution pipeline is re-initiated, based on the output from the artificial intelligence process. The maintenance solution pipeline automatically selected is reused.

Abstract: An inference analysis apparatus includes: a hypothetical logical expression designation unit that receives, in a case where any of hypothetical logical expressions constituting a hypothesis generated by inference that applies inferential knowledge to observed logical expressions is designated, the designated hypothetical logical expression; a knowledge extraction unit that extracts inferential knowledge whose consequent includes the designated hypothetical logical expression, from the inferential knowledge; and an observed logical expression extraction unit that specifies a logical expressions included in an antecedent of the extracted inferential knowledge, and extracts a observed logical expressions whose predicate is the same as the specified logical expressions, from the observed logical expressions.

Abstract: Systems, methods, and apparatuses are described herein for performing sentiment analysis on electronic communications relating to one or more image-based communications methods, such as emoji. Message data may be received. The message data may correspond to a message that is intended to be sent but has not yet been sent to an application. Using a first machine learning model, one or more subsets of the plurality of emoji may be determined. The one or more subsets of the plurality of emoji may comprise one or more different types and quantities of emoji, and may each correspond to the same or a different sentiment. Using a second machine learning model, one or more emojis may be selected from the one or more subsets. The one or more emojis selected may correspond to responses to the message.

Abstract: A reasoning engine is disclosed. Contemplated reasoning engines acquire data relating to one or more aspects of various environments. Inference engines within the reasoning engines review the acquire data, historical or current, to generate one or more hypotheses about how the aspects of the environments might be correlated, if at all. The reasoning engine can attempt to validate the hypotheses through controlling acquisition of the environment data.

Abstract: A method includes: obtaining impact values for characteristic conditions; selecting training data subsets respectively from training data sets according to the impact values; obtaining a candidate model and an evaluation value based on the training data subsets; supplementing the training data subsets according to the impact values; obtaining another candidate model and another evaluation value based on training data subsets thus supplemented; repeating the step of supplementing the training data subset, and the step of obtaining another candidate model and another evaluation value based on the training data subsets thus supplemented; and selecting one of the candidate models as a prediction model based on the evaluation values.

Abstract: A training data confirmation support device according to the present disclosure includes a label inference unit that infers inference labels that are labels corresponding to elements included in training data in which elements and correct labels corresponding to the elements are associated with each other using a model that is learned using the training data and infers labels corresponding to the elements, and an evaluation unit that generates evaluation results of the training data creators on the basis of comparison between correct labels corresponding to elements included in the training data and inference labels of the elements.

Abstract: According to one embodiment, a method, computer system, and computer program product for probabilistic inference from imprecise knowledge is provided. The embodiment may include identifying a knowledge base of one or more statements and first probability distributions corresponding to each of the one or more statements. The embodiment may also include identifying one or more queries. The embodiment may further include determining logical inferences about and second probability distributions for queries from the one or more queries or statements from the one or more statements based on information in the knowledge base.

Abstract: Aspects of the present disclosure provide techniques for classifying a trip. Embodiments include receiving, from a plurality of users, a plurality of historical trip records. Each of the plurality of historical trip records may comprise one or more historical trip attributes and historical classification information. Embodiments include training a predictive model, using the plurality of historical trip records, to classify trips based on trip records. Training the predictive model may comprise determining a plurality of hot spots based on the historical trip records, each of the plurality of hot spots comprising a region encompassing one or more locations, and associating, in the predictive model, the plurality of hot spots with historical classification information. Embodiments include receiving, from a user, a new trip record comprising a plurality of trip attributes related to a trip and using the predictive model to predict a classification for the trip based on the trip record.

Abstract: Described herein are methods and a system for analyzing the impact of multiple components with one another that support a cloud service. Events are collected in time series from the components and aggregated in a relationship tree that groups the components. Propositions as to the events are created from which a conjunctive normal form (CNF) statement is derived. The CNF statement is converted to one or more directed acyclic graphs (DAG). The DAGs are traversed to determine TRUE values used to provide remediations solutions.

Abstract: An occupancy grid probability prediction method includes: determining a dynamic particle probability based on a first probability that a dynamic particle will move from a donor cell of an occupancy grid to a recipient cell of the occupancy grid; and determining one or more predicted probabilities of at least one of the donor cell or the recipient cell based on the dynamic particle probability.

Abstract: Methods, apparatus, systems to determine a conditional probability based on audience member probability distributions for media audience measurement are disclosed. Disclosed example methods for media audience measurement include determining a first audience probability distribution for a first member of a household and determining a second audience probability distribution for a second member of the household. Disclosed example methods also include calculating probabilities for audience combinations of the first member and the second member of the household based on the first audience probability distribution and the second audience probability distribution. Disclosed example methods further include determining a household audience characteristic probability based on the calculated probabilities of the audience combinations of the household. The household audience characteristic indicates likelihoods of different possible audience compositions of the household for a media event.

Abstract: A computer-implemented method, system and computer program product for pruning quantum calculational results. The results of quantum calculations performed by a quantum circuit of a quantum computer are received, such as for each shot. A measured state of a quantum bit of a result of the quantum calculation is then compared with an expected value. If the compared measured state of the quantum bit of the result of the quantum calculation matches the expected value, then the result of the quantum calculation is not discarded. If, however, the measured state of the quantum bit of the result of the quantum calculation does not match the expected value, then the result of the quantum calculation is discarded. In this manner, the size of the full result (compilation of multiple shot results) is reduced thereby reducing the amount of classical data that needs to be stored, interpreted and transported.

Abstract: In some aspects, the techniques described herein relate to a quantum method for solving a second-order cone program (SOCP) instance, the method including: defining a Newton system for the SOCP instance by constructing matrix G and vector h based on the SOCP instance; preconditioning matrix G and vector h via row normalization to reduce a condition number of matrix G; iteratively determining u until a predetermined iteration condition is met, the iterations including: causing a quantum computing system to apply matrix G and vector h to a quantum linear system solver (QLSS) to generate a quantum state; causing the quantum computing system to perform quantum state tomography on the quantum state; and updating a value of u based on a current value of u and the output of the quantum state tomography; and determining a solution to the SOCP instance based on the updated value of u.

Abstract: A photonic element for a quantum information processing device contains a high-purity silicon layer. The high-purity silicon layer contains integrated rare-earth element (REE) dopants at a concentration of 1019 cm?3 or less. An optical transition between the lowest crystal field levels of the REE dopants integrated in the high-purity silicon layer exhibits a homogeneous linewidth of 1 MHz or less at a temperature of 4 K or less. A method for producing such a photonic element is also disclosed.

Abstract: Systems and methods for operation of a computing system to direct a search space for an optimization problem are described. One or more processors initialize an optimization algorithm, and iteratively until a termination criteria is met: receive a sample solution from the optimization algorithm, evaluate quality and feasibility of the sample solution, and where the sample solution is feasible and has the best quality so far, freeze one or more penalty parameters for a set number of iterations. Where the sample solution is not feasible or does not have the best quality so far, the one or more penalty parameters are updated based on a finite state machine, the updated one or more penalty parameters are returned to the optimization algorithm, the optimization algorithm is incremented, the termination criteria is evaluated, and when the termination criteria is met, one or more sample solutions are returned.

Abstract: Methods, systems, and apparatus for implementing a quantum circuit that moves a surface code patch of qubits. In one aspect, a method includes performing a first surface code cycle in a system of measure and data qubits. A first CNOT gate is applied to a measure qubit and a first data qubit, where the first data qubit is coupled to the measure qubit in a first direction and the first CNOT gate targets one of the measure qubits and the first data qubit. A second CNOT gate is applied to the measure qubit and the first data qubit, where the second CNOT gate targets another of the measure qubit and the first data qubit. Performing the first surface code cycle transfers information stored by the measure qubit and information stored by the first data qubit to other qubits to logically move the measure qubit and the first data qubit.

Abstract: A first quantum computing device receives a first set of computational information that reflects a utilization of computing resources of a first quantum service in a first quantum computing system that exceeds a computing resources threshold. The computing resources of the first quantum service in the first quantum computing system are altered based on the first set of computational information. The first quantum computing device then determines that the utilization of computing resources of the first quantum service in the first quantum computing system continues to exceed the computing resources threshold. The first quantum computing device causes an initiation of a copy of the first quantum service onto a second quantum computing device in a second quantum computing system.

Abstract: Information descriptive of a request to access a quantum computing resource and an intended action associated with the quantum computing resource is obtained from a quantum client device by a quantum computing system. Based at least in part on the information, a quantum computing resource access request is sent to a quantum access granting entity, wherein the quantum computing resource access request comprises information descriptive of the quantum computing resource, an intended action, and an identity of the quantum client device. A request to assign a preliminary lock to the quantum computing resource for the quantum client device is provided. Responsive to sending the quantum computing resource access request, information indicative of a decision to grant quantum computing resource access to the quantum client device is received from the quantum access granting entity.

Abstract: Systems and methods provide determination of a model script for training a first machine learning model based on input training data, determination of a metrics script for determining one or more performance metric values associated with the trained first machine learning model based on validation data, and compilation of the model script and the metrics script into an executable file.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed for managing training for models. An example apparatus includes a processor circuitry to at least obtain a request to train or retrain a model, respond to the request by preventing the train or retraining, calculate at least one performance metric, and compare performance metric corresponding to current model execution to at least one threshold performance metric.

Abstract: One or more iterations are performed. Each iteration includes calculating, for each of a number of data points that each have a label probability distribution, a label quality measure based on the label probability distribution of the data point. Each iteration includes updating the label probability distribution of each of at least one of the data points using either or both of a classification technique and a constrained clustering technique based on the data points and the label quality measure of each data point.

Abstract: Systems, methods, and computer program products are provided herein for centralized data governance within distributed component computing environments. An example method includes receiving first component metadata associated with one or more operating parameters of the first distributed computing component and second component metadata associated with one or more operating parameters of the second distributed computing component. The method includes determining, via a machine learning (ML) subsystem, a centralized governance dataset based upon the first component metadata and the second component metadata. The method further includes generating a representation of the centralized governance dataset that is accessible by one or more of the first distributed computing component or the second distributed computing component.

Abstract: A computer-implemented method for integrating text analysis of lithological descriptions with petrophysical models is described herein. The method includes preprocessing textual descriptions associated with cuttings to generate training data and generating bag-of-words vectors using the training data. The method also includes training a deep learning model to output hydrocarbon potential using the bag-of-words vectors as input and executing the trained deep learning model on unseen textual descriptions.

Abstract: Methods and systems of optimizing battery charging are disclosed. Battery state sensors are used to determine anode overpotential of a battery multiple times during multiple charge cycles. In a first phase, a reinforcement learning model (e.g., actor-critic model) is trained with rewards given throughout each charge cycle of the battery to optimize training. The reinforcement learning model can determine state-of-health characteristics of the battery over the charge cycles, and in a second phase, the reinforcement learning model is augmented accordingly. During this augmentation, the reinforcement learning model is trained with rewards given on a charge cycle-by-cycle basis, wherein rewards are given after looking at the charging optimization after the conclusion of each charge cycle.